User:BenKovitz/Salience

Essay in progress. Feel free to add/edit. Adding dissenting ideas is fine, too. Please add those at the end.

Topics to cover (not yet ordered or organized):


 * A danger for an on-line encyclopedia is clutter: filling up pages with information of low importance, distracting readers from the important facts.
 * Terminology: notability determines what's appropriate as an article topic; salience determines which facts merit inclusion in an article. "Articles should consist of salient facts on a notable topic." --Charles Matthews
 * The standard for salience in an encyclopedia article is higher than in a full-length biography, a newspaper article, or even an informal talk. We summarize and give highlights.
 * Just because information is available in verifiable sources is not a sufficient reason to include it in an article.
 * Exactly where the encyclopedic standard cuts out information is not precisely definable. It must be continually negotiated out of different editors' opinions about what is useful to readers.
 * What we share as encyclopedia editors is the common purpose of writing an encyclopedia. We might not agree on where to draw the line on salience, but we agree that we are not writing full-length biographies, newspaper articles, or other fully detailed source material.
 * A well-written biography is richly textured with detail, a job that requires a personal touch and artistic subtlety. This doesn't work on a mass wiki where coarse rules resolve editing disputes.
 * Examples of non-salient information for specific kinds of articles.
 * Counterexamples showing why we don't make strict rules about these things.
 * Ways to incorporate less-salient but still encyclopedia-worthy facts without creating clutter.
 * In biographical articles, salience derives from why the subject is notable. Birth and death dates: only of interest because the person is interesting. That a physicist got a divorce is probably not salient, but a politician's divorce might be very salient because of scandal and consequences to his or her career (explained at WP:BLP).
 * A short article is not a stub. A stub is an article that contains almost no information beyond what the topic is. For some subjects, an excellent article might contain one paragraph article, crisply stating the salient facts. Don't pad articles with trivia so they won't be "stubs". Better to remove the stub tag and give an explanation in the edit summary.
 * Examples of excellent very short articles. These should illustrate how nice it is for a reader to have the main facts about a person or topic on one page, without clutter and without having to find them inside a much larger page. Short articles are often wonderful.
 * People who are notable for one incident in their lives should (usually) have an article named for the person, not the incident, especially if the incident has no natural and obvious. Readers will likely enter the person's name into search engines, not the name chosen for the incident. An article about the person provides an orderly and natural way for "derivative salience" to go into the article (see below).
 * "Wikipedia is not paper, and nor is it a Christmas newsletter". Stuff that resembles "vanity" may do so because it lacks salience. If it looks like newslettercruft it probably should go. —Charles Matthews
 * There is more to writing an encyclopedia than collecting facts from verifiable sources. Two other things that an encyclopedist does are: organize and present those facts in a readable way that does justice to their relationships and relative importance; exercise judgement about the relative importance of facts, to decide what to include, what to leave out, and what to emphasize.
 * The existence of a fact in a primary source is, in and of itself, weak evidence of its salience. The existence of a fact in secondary sources, especially multiple secondary sources, is stronger evidence of its salience, but still not conclusive. A fact might be salient for purposes of a chapter in a book but not for Wikipedia. Salience is relative to the subject, and always a matter of human judgement. "It appeared in secondary sources" does not answer the question, "Why is it salient here, in this Wikipedia article?"
 * Why is this topic notable enough to merit an article? The basic facts that make the topic notable are always salient. Call this "primary salience". Of course, subject may be notable because of many different facts.
 * Derivative salience usually pertains to context. Birth and death dates are salient because they put a human subject into temporal context. Notice that years have their own Wikipedia articles, because years are notable.
 * Names of family members and friends, names and species of pets, sports and hobbies enjoyed, schools other than colleges attended, teachers, and towns lived in are not salient unless they have some primary salience of their own or relate to the subject in a subject-specific way. In other words, no general rules like "All mathematicians' high schools are salient." If a particular mathematician made a notable discovery using a computer at their high school, that makes the high school salient. It is part of the specific story of the notable fact. Notice that this kind of relationship that gives a fact derivative salience is almost always unique.
 * A subject's children are salient if they are notable on their own. In this case, the children should be listed, linked to their own articles. They are salient because they provide notable context for the subject. The same applies to the children of royalty, or other situations where family lineage itself makes the subject notable.
 * Another reason why family members and friends might be salient is if they contributed in some specific way to what makes the subject notable. If you list children or friends, list them in a sentence or paragraph that tells the story of the connection. Bad: "He had a sister named Abigail." Good: "When he was six years old, his sister Abigail bought him his first flute, made by the Gemeinhardt corporation"—in an article about a flutist.
 * That a fact is found in self-promoting sources, such as a company web site, is especially poor evidence of salience.
 * For any derivative salience, there should always be a clear and specific answer to the question, "Why is this salient?" If a fact is not part of what makes the subject notable, and does not directly place the subject into context of a sort that is itself notable (such as birth years), then the fact is probably cluttering the page.
 * A salient fact either answers the question "Why is this subject notable?" or relates the subject to notable elements of its context.
 * Every word in an article that does not directly relate to the reason the subject is notable carries a cost: it obscures and distracts from the facts that do explain why the subject is notable. Non-salient facts waste our readers' time and attention. To include them simply because they appear in sources is to abandon one of the primary elements of our job as encyclopedists: to exercise judgement about the relative salience of facts.
 * Facts about personal development and what a subject did later in life after the notable event occurred need to meet a lower standard of salience. For example, what became of the child TV star when he became an adult? Crimes and small accomplishments (such as patents granted) that would not be notable in themselves become salient because they answer the questions, "How did he develop talents, acquire special knowledge, and gain the social position that enabled him to do his notable event?" and "Given the promise of future accomplishments (or misdeeds) suggested by the notable event, what did the person actually do?" Social context (mentors and other personal connections that relate to the subject's notability) is salient.
 * In non-biographical articles, facts should have salience for reason of a special relation to the subject—not an ordinary relation that almost anything of its category would have. For example, in an article about a software library, it would be clutter to try to list every application that uses the library. That is best done on the web site for the software library. An application or user that is notable in itself might merit mention, especially if it is surprising and unexpected that it would use that library—for example, an accounting package not written by Microsoft being the main accounting package in use at Microsoft. The very first application to use that library would likely be salient, because that is a major milestone in the development and spread of that library. It is a major element in the "story" of the subject. The fact that some software used in nuclear power plants is written in Fortran would not be salient to an article about Fortran, unless some aspect of Fortran (perhaps commonness of bugs) related specifically to something of interest about nuclear power plants (perhaps safety), as shown in some factual source. Such a connection would show Fortran's importance in the world, as opposed to "Fortran is used in nuclear power plants."
 * It is not Wikipedia's place to give credit where credit is due. For example, an organization who funded the development of a free software library is likely not salient to the article about that library, unless there is something special to say about that organization's relation to the library, or the organization is specifically connected with it. For example, GNU merits a link in a page about any GNU software project, but Hewlett-Packard does not. Think of the bloat and clutter that would result from listing every funder of every project on Earth.
 * "If it applies to all or almost all, then it applies to none or almost none."